Questions about Chatbot and SQL structure

I've searched extensively and haven't found the answers I need.

Describe the problem/error/question

I'm unsure how to implement my idea and would appreciate any suggestions or examples.

My goal is to create a chatbot with the following capabilities:

• Access two PostgreSQL tables.
• Modify one of the tables (add or update records).
• Be scalable to accommodate 1.6k users (though not concurrently).
• The business application involves shop managers needing to report future plans to HQ. Since this is time-consuming and often relies on their internal knowledge, a chatbot would streamline the process. Managers can communicate their plans, and the chatbot will update the database accordingly.
  I'm currently connected to ChatGPT4o and Supabase (for PostgreSQL).

I've attempted the following:

• Using an AI agent with SQL capabilities, but I'm struggling to understand its functionality or find resources to guide me.
• Employing an AI agent with three PostgreSQL tools: consultDB1, consultDB2, and modifyDB2.
  • The agent has difficulty modifying DB2 correctly, so I'm hesitant to add a fourth tool for adding records.
• Utilizing an AI agent with three PostgreSQL modules: consultDB1, consultDB2, and modifyDB2.
  • I haven't been successful in getting standard results from the agent to trigger modules and then return to the chat.

I'm a beginner and undertaking this as a proof of value. Voice API integration isn't necessary at this stage; the primary requirement is database updates.

What approach would you recommend? Do you have any relevant examples?

Please share your workflow

I've tried and deleted too many configurations to recall and haven't saved them.

Information on your callin.io setup

• callin.io version: 1.67.1.
• Database (default: SQLite): PostgreSQL
• callin.io EXECUTIONS_PROCESS setting (default: own, main): Not found, so I assume "own".
• Running callin.io via: cloud
• Operating system: windows10 home